Variables and Mechanisms Affecting Response to Language Treatment
Total Page:16
File Type:pdf, Size:1020Kb
behavioral sciences Review Variables and Mechanisms Affecting Response to Language Treatment in Multilingual People with Aphasia Mira Goral 1,2,* and Aviva Lerman 3 1 Speech-Language-Hearing Sciences, Graduate Center & Lehman College, City University of New York, New York, NY 10016, USA 2 MultiLing Center for Multilingualism in Society across the Lifespan, University of Oslo, 0313 Oslo, Norway 3 Program of Communication Disorders, Hadassah Academic College, Jerusalem 9101001, Israel; [email protected] * Correspondence: [email protected] Received: 13 August 2020; Accepted: 18 September 2020; Published: 22 September 2020 Abstract: Background: Despite substantial literature exploring language treatment effects in multilingual people with aphasia (PWA), inconsistent results reported across studies make it difficult to draw firm conclusions. Methods: We highlight and illustrate variables that have been implicated in affecting cross-language treatment effects in multilingual PWA. Main contribution: We argue that opposing effects of activation and inhibition across languages, influenced by pertinent variables, such as age of language acquisition, patterns of language use, and treatment-related factors, contribute to the complex picture that has emerged from current studies of treatment in multilingual PWA. We propose a new integrated model—Treatment Effects in Aphasia in Multilingual people (the TEAM model)—to capture this complexity. Keywords: language activation; language inhibition; language treatment; cross-language generalization; stroke aphasia 1. Introduction This paper addresses the variables and mechanisms that affect efficacy of language treatment provided for multilingual people who acquire language impairments (aphasia). Our aim is to shed light on a complex phenomenon and to reflect on the mixed patterns of findings regarding cross-language response to intervention reported to date. We begin with a brief overview of aphasia generally and aphasia in multilingual people more specifically. We then review and discuss the variables that have been implicated in affecting the patterns of improvement among multilingual individuals and that contribute to levels of language activation and inhibition that have been proposed to account for the patterns observed. We do not endeavor to present an exhaustive review of the literature on intervention in multilingual people with aphasia. Instead, we set out to highlight the variables that have been discussed in this literature as factors that affect cross-language treatment generalization. Furthermore, we put forward a model that incorporates the factors that need to be considered in understanding intervention outcomes in multilingual individuals with aphasia. This model highlights the mechanisms that have been previously proposed to explain cross-language generalization and suggests how pertinent variables may affect these mechanisms. Aphasia is an acquired language disorder resulting from injury to the brain, most typically due to a stroke. People with aphasia experience difficulty with language production and comprehension. These difficulties range from mild to severe, and may affect all language modalities (e.g., speaking and reading), and one or more linguistic ability (morphology and semantics). The language impairment Behav. Sci. 2020, 10, 144; doi:10.3390/bs10090144 www.mdpi.com/journal/behavsci Behav. Sci. 2020, 10, 144 2 of 17 experienced by people with aphasia (PWA) affects their ability to communicate and thus their engagement in family, social, daily-living, and work environments. PWA typically improve their abilities over time, especially with intervention directed toward restoring the impaired abilities and also toward compensating for the deficits. However, complete recovery from aphasia is rare, e.g., [1]. Multilingual people, that is, people who use more than one language regularly [2], who acquire aphasia typically exhibit impairment in all their languages, to similar or varying degrees. Research reports on multilingual PWA have documented that a small majority of individuals experience comparable difficulties in all their languages (parallel impairment), provided they had pre-stroke high proficiency in these languages, or experience comparable impairments relative to the degree of language proficiency and use they had prior to the aphasia onset. In some cases, however, differential degrees of impairment of one or more languages are noted (non-parallel impairment) [3–5]. For example, in highly proficient bilingual people, one language can be impaired, while their other language remains relatively intact post-stroke [6]. Several variables have been implicated in determining patterns of impairment and recovery in multilingual PWA. These include variables related to the type of multilingualism, such as age of language acquisition, relative pre-stroke proficiencies of each language, and frequency of language use and exposure, as well as variables related to the stroke, such as site and size of the brain lesion and the time elapsed since the aphasia onset [7–9]. Intervention in aphasia has been shown to improve language comprehension and production and to facilitate recovery, e.g., [10]. PWA of varying severity and profiles exhibit common characteristics (e.g., word finding difficulties, termed anomia), but also a range of heterogeneous impairments and heterogeneous response to intervention. Researchers and clinicians have examined factors that influence the degree to which PWA respond to intervention and have been working toward identifying variables that can predict best response to therapy [10–12]. With multilingual PWA, in addition to the greater heterogeneity among them than among monolingual PWA (including varying degree of language abilities prior to the aphasia onset), decisions have to be made regarding which language(s) to select for the intervention. The literature on treatment effects in multilingual PWA is particularly mixed in terms of approaches used and findings reported, e.g., [13–15]. The mechanisms underlying treatment-related improvement in aphasia have been studied in recent years, but much is still not well understood [16,17]. The evidence suggests that neuroplasticity allows for reorganization of portions of the neuronal networks, and that regions adjacent to or homologous to the affected regions take on language-related processing [18,19]. Similar processes can be assumed to account for changes observed in multilingual PWA, but an additional consideration is required. Namely, expectations regarding changes associated with treatment would depend on hypotheses regarding the neuronal networks that are associated with the processing of multiple languages in the brain. Much has been debated about the underlying networks subserving multiple languages. For example, brain stimulation studies [20] and some fMRI studies [21] have demonstrated differential patterns of activation for the different languages of a multilingual person during language processing. In contrast, other neuroimaging investigations have demonstrated complete overlap in the patterns of activation during fMRI studies of multiple languages [22,23]. Moreover, Abutalebi and Green [24] argue that even when there are differences in the patterns of activation during language processing observed with neuroimaging, these can be attributed to mechanisms of activation and inhibition of the relevant languages rather than to the processing of the languages per se. Researchers who put forward mechanisms to account for the complex picture of improvement in multilingual PWA draw on what has been proposed in the study of multilingual language processing. Psycholinguistic investigations of language performance in multilingual individuals suggest that, in the presence of multiple languages, mechanisms of activation as well as mechanisms of inhibition regulate language use, depending on the communication situation [25]. During language production, words and structures from all languages of multilingual speakers are presumed to be activated. When the conversation context is multilingual, they may mix words Behav. Sci. 2020, 10, 144 3 of 17 and structures from different languages within the conversation (termed code switching) and also within a conversation turn (termed code mixing) [26,27]. If the conversation context requires the use of only one language, mechanisms of language inhibition allow the speakers to avoid language mixing. Inhibition can take place at different stages of language production—while formulating the message or selecting the sentence structure and lexical items, during the retrieval of the words, and during the selection of the word form to be produced, e.g., [28,29]. Multilingual PWA may have specific deficits in their languages, as well as impairments of the activation and inhibition mechanisms, affecting access to their languages [30]. These linguistic and mechanistic deficits can interact with a number of variables to determine the response of multilingual PWA to intervention. In Sections2–4 below, we highlight examples from the literature on treatment in multilingual PWA, illustrating variables that may account for patterns of cross-language response to language intervention. To this end, we discuss nine key variables that have dominated the literature on cross-language treatment in multilingual PWA: three multilingualism-related variables (age of language acquisition, language use and exposure, language proficiency pre stroke); two stroke-related variables (brain lesion site, time post-onset); language